[linux-2.6-block.git] / net / xfrm / xfrm_input.c

/*
 * xfrm_input.c
 *
 * Changes:
 * 	YOSHIFUJI Hideaki @USAGI
 * 		Split up af-specific portion
 *
 */

#include <linux/slab.h>
#include <linux/module.h>
#include <linux/netdevice.h>
#include <net/dst.h>
#include <net/ip.h>
#include <net/xfrm.h>
#include <net/ip_tunnels.h>
#include <net/ip6_tunnel.h>

static struct kmem_cache *secpath_cachep __read_mostly;

static DEFINE_SPINLOCK(xfrm_input_afinfo_lock);
static struct xfrm_input_afinfo const __rcu *xfrm_input_afinfo[AF_INET6 + 1];

static struct gro_cells gro_cells;
static struct net_device xfrm_napi_dev;

int xfrm_input_register_afinfo(const struct xfrm_input_afinfo *afinfo)
{
	int err = 0;

	if (WARN_ON(afinfo->family >= ARRAY_SIZE(xfrm_input_afinfo)))
		return -EAFNOSUPPORT;

	spin_lock_bh(&xfrm_input_afinfo_lock);
	if (unlikely(xfrm_input_afinfo[afinfo->family] != NULL))
		err = -EEXIST;
	else
		rcu_assign_pointer(xfrm_input_afinfo[afinfo->family], afinfo);
	spin_unlock_bh(&xfrm_input_afinfo_lock);
	return err;
}
EXPORT_SYMBOL(xfrm_input_register_afinfo);

int xfrm_input_unregister_afinfo(const struct xfrm_input_afinfo *afinfo)
{
	int err = 0;

	spin_lock_bh(&xfrm_input_afinfo_lock);
	if (likely(xfrm_input_afinfo[afinfo->family] != NULL)) {
		if (unlikely(xfrm_input_afinfo[afinfo->family] != afinfo))
			err = -EINVAL;
		else
			RCU_INIT_POINTER(xfrm_input_afinfo[afinfo->family], NULL);
	}
	spin_unlock_bh(&xfrm_input_afinfo_lock);
	synchronize_rcu();
	return err;
}
EXPORT_SYMBOL(xfrm_input_unregister_afinfo);

static const struct xfrm_input_afinfo *xfrm_input_get_afinfo(unsigned int family)
{
	const struct xfrm_input_afinfo *afinfo;

	if (WARN_ON_ONCE(family >= ARRAY_SIZE(xfrm_input_afinfo)))
		return NULL;

	rcu_read_lock();
	afinfo = rcu_dereference(xfrm_input_afinfo[family]);
	if (unlikely(!afinfo))
		rcu_read_unlock();
	return afinfo;
}

static int xfrm_rcv_cb(struct sk_buff *skb, unsigned int family, u8 protocol,
		       int err)
{
	int ret;
	const struct xfrm_input_afinfo *afinfo = xfrm_input_get_afinfo(family);

	if (!afinfo)
		return -EAFNOSUPPORT;

	ret = afinfo->callback(skb, protocol, err);
	rcu_read_unlock();

	return ret;
}

void __secpath_destroy(struct sec_path *sp)
{
	int i;
	for (i = 0; i < sp->len; i++)
		xfrm_state_put(sp->xvec[i]);
	kmem_cache_free(secpath_cachep, sp);
}
EXPORT_SYMBOL(__secpath_destroy);

struct sec_path *secpath_dup(struct sec_path *src)
{
	struct sec_path *sp;

	sp = kmem_cache_alloc(secpath_cachep, GFP_ATOMIC);
	if (!sp)
		return NULL;

	sp->len = 0;
	sp->olen = 0;

	memset(sp->ovec, 0, sizeof(sp->ovec[XFRM_MAX_OFFLOAD_DEPTH]));

	if (src) {
		int i;

		memcpy(sp, src, sizeof(*sp));
		for (i = 0; i < sp->len; i++)
			xfrm_state_hold(sp->xvec[i]);
	}
	refcount_set(&sp->refcnt, 1);
	return sp;
}
EXPORT_SYMBOL(secpath_dup);

int secpath_set(struct sk_buff *skb)
{
	struct sec_path *sp;

	/* Allocate new secpath or COW existing one. */
	if (!skb->sp || refcount_read(&skb->sp->refcnt) != 1) {
		sp = secpath_dup(skb->sp);
		if (!sp)
			return -ENOMEM;

		if (skb->sp)
			secpath_put(skb->sp);
		skb->sp = sp;
	}
	return 0;
}
EXPORT_SYMBOL(secpath_set);

/* Fetch spi and seq from ipsec header */

int xfrm_parse_spi(struct sk_buff *skb, u8 nexthdr, __be32 *spi, __be32 *seq)
{
	int offset, offset_seq;
	int hlen;

	switch (nexthdr) {
	case IPPROTO_AH:
		hlen = sizeof(struct ip_auth_hdr);
		offset = offsetof(struct ip_auth_hdr, spi);
		offset_seq = offsetof(struct ip_auth_hdr, seq_no);
		break;
	case IPPROTO_ESP:
		hlen = sizeof(struct ip_esp_hdr);
		offset = offsetof(struct ip_esp_hdr, spi);
		offset_seq = offsetof(struct ip_esp_hdr, seq_no);
		break;
	case IPPROTO_COMP:
		if (!pskb_may_pull(skb, sizeof(struct ip_comp_hdr)))
			return -EINVAL;
		*spi = htonl(ntohs(*(__be16 *)(skb_transport_header(skb) + 2)));
		*seq = 0;
		return 0;
	default:
		return 1;
	}

	if (!pskb_may_pull(skb, hlen))
		return -EINVAL;

	*spi = *(__be32 *)(skb_transport_header(skb) + offset);
	*seq = *(__be32 *)(skb_transport_header(skb) + offset_seq);
	return 0;
}
EXPORT_SYMBOL(xfrm_parse_spi);

int xfrm_prepare_input(struct xfrm_state *x, struct sk_buff *skb)
{
	struct xfrm_mode *inner_mode = x->inner_mode;
	int err;

	err = x->outer_mode->afinfo->extract_input(x, skb);
	if (err)
		return err;

	if (x->sel.family == AF_UNSPEC) {
		inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
		if (inner_mode == NULL)
			return -EAFNOSUPPORT;
	}

	skb->protocol = inner_mode->afinfo->eth_proto;
	return inner_mode->input2(x, skb);
}
EXPORT_SYMBOL(xfrm_prepare_input);

int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
{
	struct net *net = dev_net(skb->dev);
	int err;
	__be32 seq;
	__be32 seq_hi;
	struct xfrm_state *x = NULL;
	xfrm_address_t *daddr;
	struct xfrm_mode *inner_mode;
	u32 mark = skb->mark;
	unsigned int family;
	int decaps = 0;
	int async = 0;
	bool xfrm_gro = false;
	bool crypto_done = false;
	struct xfrm_offload *xo = xfrm_offload(skb);

	if (encap_type < 0) {
		x = xfrm_input_state(skb);
		family = x->outer_mode->afinfo->family;

		/* An encap_type of -1 indicates async resumption. */
		if (encap_type == -1) {
			async = 1;
			seq = XFRM_SKB_CB(skb)->seq.input.low;
			goto resume;
		}

		/* encap_type < -1 indicates a GRO call. */
		encap_type = 0;
		seq = XFRM_SPI_SKB_CB(skb)->seq;

		if (xo && (xo->flags & CRYPTO_DONE)) {
			crypto_done = true;
			x = xfrm_input_state(skb);
			family = XFRM_SPI_SKB_CB(skb)->family;

			if (!(xo->status & CRYPTO_SUCCESS)) {
				if (xo->status &
				    (CRYPTO_TRANSPORT_AH_AUTH_FAILED |
				     CRYPTO_TRANSPORT_ESP_AUTH_FAILED |
				     CRYPTO_TUNNEL_AH_AUTH_FAILED |
				     CRYPTO_TUNNEL_ESP_AUTH_FAILED)) {

					xfrm_audit_state_icvfail(x, skb,
								 x->type->proto);
					x->stats.integrity_failed++;
					XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
					goto drop;
				}

				if (xo->status & CRYPTO_INVALID_PROTOCOL) {
					XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
					goto drop;
				}

				XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
				goto drop;
			}

			if ((err = xfrm_parse_spi(skb, nexthdr, &spi, &seq)) != 0) {
				XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
				goto drop;
			}
		}

		goto lock;
	}

	daddr = (xfrm_address_t *)(skb_network_header(skb) +
				   XFRM_SPI_SKB_CB(skb)->daddroff);
	family = XFRM_SPI_SKB_CB(skb)->family;

	/* if tunnel is present override skb->mark value with tunnel i_key */
	switch (family) {
	case AF_INET:
		if (XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip4)
			mark = be32_to_cpu(XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip4->parms.i_key);
		break;
	case AF_INET6:
		if (XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6)
			mark = be32_to_cpu(XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6->parms.i_key);
		break;
	}

	err = secpath_set(skb);
	if (err) {
		XFRM_INC_STATS(net, LINUX_MIB_XFRMINERROR);
		goto drop;
	}

	seq = 0;
	if (!spi && (err = xfrm_parse_spi(skb, nexthdr, &spi, &seq)) != 0) {
		XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
		goto drop;
	}

	do {
		if (skb->sp->len == XFRM_MAX_DEPTH) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
			goto drop;
		}

		x = xfrm_state_lookup(net, mark, daddr, spi, nexthdr, family);
		if (x == NULL) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINNOSTATES);
			xfrm_audit_state_notfound(skb, family, spi, seq);
			goto drop;
		}

		skb->sp->xvec[skb->sp->len++] = x;

lock:
		spin_lock(&x->lock);

		if (unlikely(x->km.state != XFRM_STATE_VALID)) {
			if (x->km.state == XFRM_STATE_ACQ)
				XFRM_INC_STATS(net, LINUX_MIB_XFRMACQUIREERROR);
			else
				XFRM_INC_STATS(net,
					       LINUX_MIB_XFRMINSTATEINVALID);
			goto drop_unlock;
		}

		if ((x->encap ? x->encap->encap_type : 0) != encap_type) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMISMATCH);
			goto drop_unlock;
		}

		if (x->repl->check(x, skb, seq)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
			goto drop_unlock;
		}

		if (xfrm_state_check_expire(x)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEEXPIRED);
			goto drop_unlock;
		}

		spin_unlock(&x->lock);

		if (xfrm_tunnel_check(skb, x, family)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
			goto drop;
		}

		seq_hi = htonl(xfrm_replay_seqhi(x, seq));

		XFRM_SKB_CB(skb)->seq.input.low = seq;
		XFRM_SKB_CB(skb)->seq.input.hi = seq_hi;

		skb_dst_force(skb);
		dev_hold(skb->dev);

		if (crypto_done)
			nexthdr = x->type_offload->input_tail(x, skb);
		else
			nexthdr = x->type->input(x, skb);

		if (nexthdr == -EINPROGRESS)
			return 0;
resume:
		dev_put(skb->dev);

		spin_lock(&x->lock);
		if (nexthdr <= 0) {
			if (nexthdr == -EBADMSG) {
				xfrm_audit_state_icvfail(x, skb,
							 x->type->proto);
				x->stats.integrity_failed++;
			}
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
			goto drop_unlock;
		}

		/* only the first xfrm gets the encap type */
		encap_type = 0;

		if (async && x->repl->recheck(x, skb, seq)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
			goto drop_unlock;
		}

		x->repl->advance(x, seq);

		x->curlft.bytes += skb->len;
		x->curlft.packets++;

		spin_unlock(&x->lock);

		XFRM_MODE_SKB_CB(skb)->protocol = nexthdr;

		inner_mode = x->inner_mode;

		if (x->sel.family == AF_UNSPEC) {
			inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
			if (inner_mode == NULL) {
				XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
				goto drop;
			}
		}

		if (inner_mode->input(x, skb)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
			goto drop;
		}

		if (x->outer_mode->flags & XFRM_MODE_FLAG_TUNNEL) {
			decaps = 1;
			break;
		}

		/*
		 * We need the inner address.  However, we only get here for
		 * transport mode so the outer address is identical.
		 */
		daddr = &x->id.daddr;
		family = x->outer_mode->afinfo->family;

		err = xfrm_parse_spi(skb, nexthdr, &spi, &seq);
		if (err < 0) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
			goto drop;
		}
	} while (!err);

	err = xfrm_rcv_cb(skb, family, x->type->proto, 0);
	if (err)
		goto drop;

	nf_reset(skb);

	if (decaps) {
		skb->sp->olen = 0;
		skb_dst_drop(skb);
		gro_cells_receive(&gro_cells, skb);
		return 0;
	} else {
		xo = xfrm_offload(skb);
		if (xo)
			xfrm_gro = xo->flags & XFRM_GRO;

		err = x->inner_mode->afinfo->transport_finish(skb, xfrm_gro || async);
		if (xfrm_gro) {
			skb->sp->olen = 0;
			skb_dst_drop(skb);
			gro_cells_receive(&gro_cells, skb);
			return err;
		}

		return err;
	}

drop_unlock:
	spin_unlock(&x->lock);
drop:
	xfrm_rcv_cb(skb, family, x && x->type ? x->type->proto : nexthdr, -1);
	kfree_skb(skb);
	return 0;
}
EXPORT_SYMBOL(xfrm_input);

int xfrm_input_resume(struct sk_buff *skb, int nexthdr)
{
	return xfrm_input(skb, nexthdr, 0, -1);
}
EXPORT_SYMBOL(xfrm_input_resume);

void __init xfrm_input_init(void)
{
	int err;

	init_dummy_netdev(&xfrm_napi_dev);
	err = gro_cells_init(&gro_cells, &xfrm_napi_dev);
	if (err)
		gro_cells.cells = NULL;

	secpath_cachep = kmem_cache_create("secpath_cache",
					   sizeof(struct sec_path),
					   0, SLAB_HWCACHE_ALIGN|SLAB_PANIC,
					   NULL);
}
Commit	Line	Data
1da177e4 LT	1	/*
	2	* xfrm_input.c
	3	*
	4	* Changes:
	5	* YOSHIFUJI Hideaki @USAGI
	6	* Split up af-specific portion
a716c119	7	*
1da177e4 LT	8	*/
	9
	10	#include <linux/slab.h>
	11	#include <linux/module.h>
716062fd HX	12	#include <linux/netdevice.h>
716062fd HX	13	#include <net/dst.h>
1da177e4 LT	14	#include <net/ip.h>
1da177e4 LT	15	#include <net/xfrm.h>
049f8e2e AD	16	#include <net/ip_tunnels.h>
049f8e2e AD	17	#include <net/ip6_tunnel.h>
1da177e4	18
e18b890b	19	static struct kmem_cache *secpath_cachep __read_mostly;
1da177e4	20
2f32b51b	21	static DEFINE_SPINLOCK(xfrm_input_afinfo_lock);
960fdfde	22	static struct xfrm_input_afinfo const __rcu *xfrm_input_afinfo[AF_INET6 + 1];
2f32b51b	23
1995876a SK	24	static struct gro_cells gro_cells;
	25	static struct net_device xfrm_napi_dev;
	26
960fdfde	27	int xfrm_input_register_afinfo(const struct xfrm_input_afinfo *afinfo)
2f32b51b SK	28	{
	29	int err = 0;
	30
960fdfde	31	if (WARN_ON(afinfo->family >= ARRAY_SIZE(xfrm_input_afinfo)))
2f32b51b	32	return -EAFNOSUPPORT;
960fdfde	33
2f32b51b SK	34	spin_lock_bh(&xfrm_input_afinfo_lock);
2f32b51b SK	35	if (unlikely(xfrm_input_afinfo[afinfo->family] != NULL))
f31e8d4f	36	err = -EEXIST;
2f32b51b SK	37	else
	38	rcu_assign_pointer(xfrm_input_afinfo[afinfo->family], afinfo);
	39	spin_unlock_bh(&xfrm_input_afinfo_lock);
	40	return err;
	41	}
	42	EXPORT_SYMBOL(xfrm_input_register_afinfo);
	43
960fdfde	44	int xfrm_input_unregister_afinfo(const struct xfrm_input_afinfo *afinfo)
2f32b51b SK	45	{
	46	int err = 0;
	47
2f32b51b SK	48	spin_lock_bh(&xfrm_input_afinfo_lock);
	49	if (likely(xfrm_input_afinfo[afinfo->family] != NULL)) {
	50	if (unlikely(xfrm_input_afinfo[afinfo->family] != afinfo))
	51	err = -EINVAL;
	52	else
	53	RCU_INIT_POINTER(xfrm_input_afinfo[afinfo->family], NULL);
	54	}
	55	spin_unlock_bh(&xfrm_input_afinfo_lock);
	56	synchronize_rcu();
	57	return err;
	58	}
	59	EXPORT_SYMBOL(xfrm_input_unregister_afinfo);
	60
960fdfde	61	static const struct xfrm_input_afinfo *xfrm_input_get_afinfo(unsigned int family)
2f32b51b	62	{
960fdfde	63	const struct xfrm_input_afinfo *afinfo;
2f32b51b	64
960fdfde	65	if (WARN_ON_ONCE(family >= ARRAY_SIZE(xfrm_input_afinfo)))
2f32b51b	66	return NULL;
960fdfde	67
2f32b51b SK	68	rcu_read_lock();
	69	afinfo = rcu_dereference(xfrm_input_afinfo[family]);
	70	if (unlikely(!afinfo))
	71	rcu_read_unlock();
	72	return afinfo;
	73	}
	74
2f32b51b SK	75	static int xfrm_rcv_cb(struct sk_buff *skb, unsigned int family, u8 protocol,
	76	int err)
	77	{
	78	int ret;
960fdfde	79	const struct xfrm_input_afinfo *afinfo = xfrm_input_get_afinfo(family);
2f32b51b SK	80
	81	if (!afinfo)
	82	return -EAFNOSUPPORT;
	83
	84	ret = afinfo->callback(skb, protocol, err);
960fdfde	85	rcu_read_unlock();
2f32b51b SK	86
	87	return ret;
	88	}
	89
1da177e4 LT	90	void __secpath_destroy(struct sec_path *sp)
	91	{
	92	int i;
	93	for (i = 0; i < sp->len; i++)
dbe5b4aa	94	xfrm_state_put(sp->xvec[i]);
1da177e4 LT	95	kmem_cache_free(secpath_cachep, sp);
	96	}
	97	EXPORT_SYMBOL(__secpath_destroy);
	98
	99	struct sec_path secpath_dup(struct sec_path src)
	100	{
	101	struct sec_path *sp;
	102
54e6ecb2	103	sp = kmem_cache_alloc(secpath_cachep, GFP_ATOMIC);
1da177e4 LT	104	if (!sp)
	105	return NULL;
	106
	107	sp->len = 0;
54ef207a SK	108	sp->olen = 0;
54ef207a SK	109
d77e38e6 SK	110	memset(sp->ovec, 0, sizeof(sp->ovec[XFRM_MAX_OFFLOAD_DEPTH]));
d77e38e6 SK	111
1da177e4 LT	112	if (src) {
	113	int i;
	114
	115	memcpy(sp, src, sizeof(*sp));
	116	for (i = 0; i < sp->len; i++)
dbe5b4aa	117	xfrm_state_hold(sp->xvec[i]);
1da177e4	118	}
55eabed6	119	refcount_set(&sp->refcnt, 1);
1da177e4 LT	120	return sp;
	121	}
	122	EXPORT_SYMBOL(secpath_dup);
	123
b0fcee82 SK	124	int secpath_set(struct sk_buff *skb)
	125	{
	126	struct sec_path *sp;
	127
	128	/* Allocate new secpath or COW existing one. */
55eabed6	129	if (!skb->sp \|\| refcount_read(&skb->sp->refcnt) != 1) {
b0fcee82 SK	130	sp = secpath_dup(skb->sp);
	131	if (!sp)
	132	return -ENOMEM;
	133
	134	if (skb->sp)
	135	secpath_put(skb->sp);
	136	skb->sp = sp;
	137	}
	138	return 0;
	139	}
	140	EXPORT_SYMBOL(secpath_set);
	141
1da177e4 LT	142	/* Fetch spi and seq from ipsec header */
1da177e4 LT	143
6067b2ba	144	int xfrm_parse_spi(struct sk_buff skb, u8 nexthdr, __be32 spi, __be32 *seq)
1da177e4 LT	145	{
1da177e4 LT	146	int offset, offset_seq;
44072500	147	int hlen;
1da177e4 LT	148
	149	switch (nexthdr) {
	150	case IPPROTO_AH:
44072500	151	hlen = sizeof(struct ip_auth_hdr);
1da177e4 LT	152	offset = offsetof(struct ip_auth_hdr, spi);
	153	offset_seq = offsetof(struct ip_auth_hdr, seq_no);
	154	break;
	155	case IPPROTO_ESP:
44072500	156	hlen = sizeof(struct ip_esp_hdr);
1da177e4 LT	157	offset = offsetof(struct ip_esp_hdr, spi);
	158	offset_seq = offsetof(struct ip_esp_hdr, seq_no);
	159	break;
	160	case IPPROTO_COMP:
	161	if (!pskb_may_pull(skb, sizeof(struct ip_comp_hdr)))
	162	return -EINVAL;
3e94c2dc	163	spi = htonl(ntohs((__be16 *)(skb_transport_header(skb) + 2)));
1da177e4 LT	164	*seq = 0;
	165	return 0;
	166	default:
	167	return 1;
	168	}
	169
44072500	170	if (!pskb_may_pull(skb, hlen))
1da177e4 LT	171	return -EINVAL;
1da177e4 LT	172
3e94c2dc WC	173	spi = (__be32 *)(skb_transport_header(skb) + offset);
3e94c2dc WC	174	seq = (__be32 *)(skb_transport_header(skb) + offset_seq);
1da177e4 LT	175	return 0;
1da177e4 LT	176	}
1e295370	177	EXPORT_SYMBOL(xfrm_parse_spi);
1da177e4	178
227620e2 HX	179	int xfrm_prepare_input(struct xfrm_state x, struct sk_buff skb)
227620e2 HX	180	{
df9dcb45	181	struct xfrm_mode *inner_mode = x->inner_mode;
227620e2 HX	182	int err;
	183
	184	err = x->outer_mode->afinfo->extract_input(x, skb);
	185	if (err)
	186	return err;
	187
df9dcb45 KM	188	if (x->sel.family == AF_UNSPEC) {
	189	inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
	190	if (inner_mode == NULL)
	191	return -EAFNOSUPPORT;
	192	}
	193
	194	skb->protocol = inner_mode->afinfo->eth_proto;
	195	return inner_mode->input2(x, skb);
227620e2 HX	196	}
	197	EXPORT_SYMBOL(xfrm_prepare_input);
	198
716062fd HX	199	int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
716062fd HX	200	{
bd235e3c	201	struct net *net = dev_net(skb->dev);
716062fd HX	202	int err;
716062fd HX	203	__be32 seq;
2cd08467	204	__be32 seq_hi;
3328715e	205	struct xfrm_state *x = NULL;
1bf06cd2	206	xfrm_address_t *daddr;
df9dcb45	207	struct xfrm_mode *inner_mode;
049f8e2e	208	u32 mark = skb->mark;
2fcb45b6	209	unsigned int family;
716062fd	210	int decaps = 0;
1bf06cd2	211	int async = 0;
7785bba2	212	bool xfrm_gro = false;
d77e38e6 SK	213	bool crypto_done = false;
d77e38e6 SK	214	struct xfrm_offload *xo = xfrm_offload(skb);
1bf06cd2	215
1bf06cd2	216	if (encap_type < 0) {
00501121	217	x = xfrm_input_state(skb);
3328715e	218	family = x->outer_mode->afinfo->family;
7785bba2 SK	219
	220	/* An encap_type of -1 indicates async resumption. */
	221	if (encap_type == -1) {
	222	async = 1;
	223	seq = XFRM_SKB_CB(skb)->seq.input.low;
	224	goto resume;
	225	}
bcd1f8a4	226
7785bba2 SK	227	/* encap_type < -1 indicates a GRO call. */
	228	encap_type = 0;
	229	seq = XFRM_SPI_SKB_CB(skb)->seq;
716062fd	230
bcd1f8a4 SK	231	if (xo && (xo->flags & CRYPTO_DONE)) {
	232	crypto_done = true;
	233	x = xfrm_input_state(skb);
	234	family = XFRM_SPI_SKB_CB(skb)->family;
	235
	236	if (!(xo->status & CRYPTO_SUCCESS)) {
	237	if (xo->status &
	238	(CRYPTO_TRANSPORT_AH_AUTH_FAILED \|
	239	CRYPTO_TRANSPORT_ESP_AUTH_FAILED \|
	240	CRYPTO_TUNNEL_AH_AUTH_FAILED \|
	241	CRYPTO_TUNNEL_ESP_AUTH_FAILED)) {
	242
	243	xfrm_audit_state_icvfail(x, skb,
	244	x->type->proto);
	245	x->stats.integrity_failed++;
	246	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
	247	goto drop;
	248	}
	249
47ebcc0b YK	250	if (xo->status & CRYPTO_INVALID_PROTOCOL) {
	251	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
	252	goto drop;
	253	}
	254
bcd1f8a4	255	XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
d77e38e6 SK	256	goto drop;
	257	}
	258
bcd1f8a4 SK	259	if ((err = xfrm_parse_spi(skb, nexthdr, &spi, &seq)) != 0) {
	260	XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
	261	goto drop;
	262	}
d77e38e6 SK	263	}
d77e38e6 SK	264
7785bba2	265	goto lock;
1bf06cd2	266	}
716062fd	267
3328715e SK	268	daddr = (xfrm_address_t *)(skb_network_header(skb) +
	269	XFRM_SPI_SKB_CB(skb)->daddroff);
	270	family = XFRM_SPI_SKB_CB(skb)->family;
	271
049f8e2e	272	/* if tunnel is present override skb->mark value with tunnel i_key */
1625f452 AK	273	switch (family) {
	274	case AF_INET:
	275	if (XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip4)
049f8e2e	276	mark = be32_to_cpu(XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip4->parms.i_key);
1625f452 AK	277	break;
	278	case AF_INET6:
	279	if (XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6)
049f8e2e	280	mark = be32_to_cpu(XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6->parms.i_key);
1625f452	281	break;
049f8e2e AD	282	}
049f8e2e AD	283
b0fcee82 SK	284	err = secpath_set(skb);
	285	if (err) {
	286	XFRM_INC_STATS(net, LINUX_MIB_XFRMINERROR);
	287	goto drop;
b2aa5e9d HX	288	}
b2aa5e9d HX	289
716062fd	290	seq = 0;
0aa64774	291	if (!spi && (err = xfrm_parse_spi(skb, nexthdr, &spi, &seq)) != 0) {
59c9940e	292	XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
716062fd	293	goto drop;
0aa64774	294	}
716062fd HX	295
716062fd HX	296	do {
0aa64774	297	if (skb->sp->len == XFRM_MAX_DEPTH) {
59c9940e	298	XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
716062fd	299	goto drop;
0aa64774	300	}
716062fd	301
049f8e2e	302	x = xfrm_state_lookup(net, mark, daddr, spi, nexthdr, family);
0aa64774	303	if (x == NULL) {
59c9940e	304	XFRM_INC_STATS(net, LINUX_MIB_XFRMINNOSTATES);
afeb14b4	305	xfrm_audit_state_notfound(skb, family, spi, seq);
716062fd	306	goto drop;
0aa64774	307	}
716062fd	308
b2aa5e9d HX	309	skb->sp->xvec[skb->sp->len++] = x;
b2aa5e9d HX	310
7785bba2	311	lock:
716062fd	312	spin_lock(&x->lock);
4c4d41f2	313
0aa64774	314	if (unlikely(x->km.state != XFRM_STATE_VALID)) {
dc0565ce LR	315	if (x->km.state == XFRM_STATE_ACQ)
	316	XFRM_INC_STATS(net, LINUX_MIB_XFRMACQUIREERROR);
	317	else
	318	XFRM_INC_STATS(net,
	319	LINUX_MIB_XFRMINSTATEINVALID);
716062fd	320	goto drop_unlock;
0aa64774	321	}
716062fd	322
3de77cf2 DM	323	if ((x->encap ? x->encap->encap_type : 0) != encap_type) {
	324	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMISMATCH);
	325	goto drop_unlock;
	326	}
	327
36ae0148	328	if (x->repl->check(x, skb, seq)) {
59c9940e	329	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
716062fd	330	goto drop_unlock;
0aa64774	331	}
716062fd	332
0aa64774	333	if (xfrm_state_check_expire(x)) {
59c9940e	334	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEEXPIRED);
716062fd	335	goto drop_unlock;
0aa64774	336	}
716062fd	337
0ebea8ef HX	338	spin_unlock(&x->lock);
0ebea8ef HX	339
68c11e98 AD	340	if (xfrm_tunnel_check(skb, x, family)) {
	341	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
	342	goto drop;
	343	}
	344
2cd08467 SK	345	seq_hi = htonl(xfrm_replay_seqhi(x, seq));
2cd08467 SK	346
1ce3644a	347	XFRM_SKB_CB(skb)->seq.input.low = seq;
2cd08467	348	XFRM_SKB_CB(skb)->seq.input.hi = seq_hi;
1bf06cd2	349
3bc07321	350	skb_dst_force(skb);
071d36bf	351	dev_hold(skb->dev);
3bc07321	352
d77e38e6 SK	353	if (crypto_done)
	354	nexthdr = x->type_offload->input_tail(x, skb);
	355	else
	356	nexthdr = x->type->input(x, skb);
0ebea8ef	357
1bf06cd2 HX	358	if (nexthdr == -EINPROGRESS)
1bf06cd2 HX	359	return 0;
1bf06cd2	360	resume:
071d36bf	361	dev_put(skb->dev);
071d36bf	362
0ebea8ef	363	spin_lock(&x->lock);
668dc8af	364	if (nexthdr <= 0) {
9dd3245a HX	365	if (nexthdr == -EBADMSG) {
	366	xfrm_audit_state_icvfail(x, skb,
	367	x->type->proto);
668dc8af	368	x->stats.integrity_failed++;
9dd3245a	369	}
59c9940e	370	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
716062fd	371	goto drop_unlock;
668dc8af	372	}
716062fd	373
716062fd HX	374	/* only the first xfrm gets the encap type */
	375	encap_type = 0;
	376
3b59df46	377	if (async && x->repl->recheck(x, skb, seq)) {
bcf66bf5 SK	378	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
	379	goto drop_unlock;
	380	}
	381
9fdc4883	382	x->repl->advance(x, seq);
716062fd HX	383
	384	x->curlft.bytes += skb->len;
	385	x->curlft.packets++;
	386
	387	spin_unlock(&x->lock);
	388
60d5fcfb HX	389	XFRM_MODE_SKB_CB(skb)->protocol = nexthdr;
60d5fcfb HX	390
df9dcb45 KM	391	inner_mode = x->inner_mode;
	392
	393	if (x->sel.family == AF_UNSPEC) {
	394	inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
cb866e32 SK	395	if (inner_mode == NULL) {
cb866e32 SK	396	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
df9dcb45	397	goto drop;
cb866e32	398	}
df9dcb45 KM	399	}
	400
	401	if (inner_mode->input(x, skb)) {
59c9940e	402	XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
716062fd	403	goto drop;
0aa64774	404	}
716062fd HX	405
	406	if (x->outer_mode->flags & XFRM_MODE_FLAG_TUNNEL) {
	407	decaps = 1;
	408	break;
	409	}
	410
1bf06cd2 HX	411	/*
	412	* We need the inner address. However, we only get here for
	413	* transport mode so the outer address is identical.
	414	*/
	415	daddr = &x->id.daddr;
2fcb45b6	416	family = x->outer_mode->afinfo->family;
1bf06cd2	417
716062fd	418	err = xfrm_parse_spi(skb, nexthdr, &spi, &seq);
0aa64774	419	if (err < 0) {
59c9940e	420	XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
716062fd	421	goto drop;
0aa64774	422	}
716062fd HX	423	} while (!err);
716062fd HX	424
3328715e SK	425	err = xfrm_rcv_cb(skb, family, x->type->proto, 0);
	426	if (err)
	427	goto drop;
	428
716062fd HX	429	nf_reset(skb);
	430
	431	if (decaps) {
7e9e9202	432	skb->sp->olen = 0;
adf30907	433	skb_dst_drop(skb);
1995876a	434	gro_cells_receive(&gro_cells, skb);
716062fd HX	435	return 0;
716062fd HX	436	} else {
7785bba2 SK	437	xo = xfrm_offload(skb);
	438	if (xo)
	439	xfrm_gro = xo->flags & XFRM_GRO;
	440
cfcf99f9	441	err = x->inner_mode->afinfo->transport_finish(skb, xfrm_gro \|\| async);
7785bba2	442	if (xfrm_gro) {
7e9e9202	443	skb->sp->olen = 0;
7785bba2 SK	444	skb_dst_drop(skb);
	445	gro_cells_receive(&gro_cells, skb);
	446	return err;
	447	}
	448
	449	return err;
716062fd HX	450	}
	451
	452	drop_unlock:
	453	spin_unlock(&x->lock);
716062fd	454	drop:
3328715e	455	xfrm_rcv_cb(skb, family, x && x->type ? x->type->proto : nexthdr, -1);
716062fd HX	456	kfree_skb(skb);
	457	return 0;
	458	}
	459	EXPORT_SYMBOL(xfrm_input);
	460
1bf06cd2 HX	461	int xfrm_input_resume(struct sk_buff *skb, int nexthdr)
	462	{
	463	return xfrm_input(skb, nexthdr, 0, -1);
	464	}
	465	EXPORT_SYMBOL(xfrm_input_resume);
	466
1da177e4 LT	467	void __init xfrm_input_init(void)
1da177e4 LT	468	{
1995876a SK	469	int err;
	470
	471	init_dummy_netdev(&xfrm_napi_dev);
	472	err = gro_cells_init(&gro_cells, &xfrm_napi_dev);
	473	if (err)
	474	gro_cells.cells = NULL;
	475
1da177e4 LT	476	secpath_cachep = kmem_cache_create("secpath_cache",
1da177e4 LT	477	sizeof(struct sec_path),
e5d679f3	478	0, SLAB_HWCACHE_ALIGN\|SLAB_PANIC,
20c2df83	479	NULL);
1da177e4	480	}